We propose to conduct and analyze a randomized-block, double-blind clinical trial to test the efficacy of 2 chronotherapeutic options to improve intraday diabetic instability of adult insulin dependent diabetics who are in a 2 injection per day insulin schedule. The proposed treatments consist of changes in the timing of insulin injection. Total doses per injection of insulin and type of insulin will remain the same. The treatments are: 1) Conventional (control): injections at 0800 and 1700 hours; 2) Earlier am insulin injection: injections at 0600 and 1700 hours; 3) Later pm injection: injections at 0800 and 2300 hours. Each group will have 16 patients, making a total of 48 patients. The rationale behind the study objectives, that is, to improve intraday blood glucose variation is based on the previous knowledge that type I diabetics exhibit a rhythm in blood glucose characterized by a rise in blood glucose level that becomes evident from about 0500 to 0800 hours, and continues after breakfast until about 0900 to 1000 hours. This period of glucose rise is also characterized by a diminished insulin sensitivity, as demonstrated by studies with the artificial pancreas. If more insulin can be provided during this period of diminished insulin sensitivity, the rise in glucose may be attenuated or even avoided, and, consequently, the post-breakfast blood glucose peak may be also decreased. As this postprandial peak has been consistently reported to be the daily blood glucose maximum, range (maximum minus minimum blood glucose), a measure of intraday diabetic instability, may be expected to be attenuated. This trial will only test the short-term benefits of the proposed treatments. With the knowledge gained by this study, the intermediate and long term benefits may be evaluated.